abnormalities in neuromuscular junction structure …s copies/cv1357.pdf · 2010. 9. 3. ·...

AAT

MPAa

Mb

c

ld

Aajmceefidmas(sltfaifinifr

Km

IcStcba

*7EAEPm

Neuroscience 148 (2007) 700–711

0d

BNORMALITIES IN NEUROMUSCULAR JUNCTION STRUCTUREND SKELETAL MUSCLE FUNCTION IN MICE LACKING

HE P2X2 NUCLEOTIDE RECEPTOR

(Edjprrppt

tsvPa12rbocetm2cic1eii

ogs

A

Ppa

I

Fa

. RYTEN,a R. KOSHI,b G. E. KNIGHT,a M. TURMAINE,c

. DUNN,a D. A. COCKAYNE,d A. P. W. FORDd

ND G. BURNSTOCKa*

Autonomic Neuroscience Centre, Royal Free and University Collegeedical School, Rowland Hill Street, London NW3 2PF, UK

Department of Anatomy, Christian Medical College, Vellore, India

Department of Anatomy and Developmental Biology, University Col-ege London, London WC1E 6BT, UK

The Neurobiology Unit, Roche Bioscience, Palo Alto, USA

bstract—ATP is co-released in significant quantities withcetylcholine from motor neurons at skeletal neuromuscularunctions (NMJ). However, the role of this neurotransmitter in

uscle function remains unclear. The P2X2 ion channel re-eptor subunit is expressed during development of the skel-tal NMJ, but not in adult muscle fibers, although it is re-xpressed during muscle fiber regeneration. Using mice de-cient for the P2X2 receptor subunit for ATP (P2X2

�/�), weemonstrate a role for purinergic signaling in NMJ develop-ent. Whereas control NMJs were characterized by precise

pposition of pre-synaptic motor nerve terminals and post-ynaptic junctional folds rich in acetylcholine receptorsAChRs), NMJs in P2X2

�/� mice were disorganized: misappo-ition of nerve terminals and post-synaptic AChR expressionocalization was common; the density of post-synaptic junc-ional folds was reduced; and there was increased end-plateragmentation. These changes in NMJ structure were associ-ted with muscle fiber atrophy. In addition there was anncrease in the proportion of fast type muscle fibers. Thesendings demonstrate a role for P2X2 receptor-mediated sig-aling in NMJ formation and suggest that purinergic signal-

ng may play an as yet largely unrecognized part in synapseormation. © 2007 IBRO. Published by Elsevier Ltd. All rightseserved.

ey words: ATP, acetylcholine, knockout, mouse, nerve ter-inal, synapse.

t is well established that ATP is co-released with acetyl-holine (ACh) from motor nerve terminals (Redman andilinsky, 1994; Redman and Silinsky, 1996). While ACh is

he transmitter that mediates via nicotinic receptors muscleontraction in mature animals, during early developmentoth ATP, acting via P2X ion channel receptors (Ralevicnd Burnstock, 1998) and ACh mediate muscle responses

Corresponding author. Tel: �44-0-20-7830-2948; fax: �44-0-20-830-2949.-mail address: [email protected] (G. Burnstock).bbreviations: ACh, acetylcholine; AChR, acetylcholine receptor;, embryonic day; NMJ, neuromuscular junction; P, postnatal day;

�/�

aBS, phosphate-buffered saline; P2X2 , P2X2 receptor knockoutouse.

306-4522/07$30.00�0.00 © 2007 IBRO. Published by Elsevier Ltd. All rights reseroi:10.1016/j.neuroscience.2007.06.050

700

Kolb and Wakelam, 1983; Henning, 1997; Heilbronn andriksson, 1998). P2X2 receptor subunits are expresseduring early postnatal development, when neuromuscular

unction (NMJ) maturation and patterning occur, but disap-ear in the adult (Ryten et al., 2001). Furthermore, thiseceptor is re-expressed in the later stages of muscleegeneration in the mdx mouse model of muscular dystro-hy (Ryten et al., 2004; Jiang et al., 2005). Also directostjunctional responses to ATP reappear after denerva-ion of chick skeletal muscle (Wells et al., 1995).

Electrophysiology, immunohistochemistry and reverseranscriptase polymerase chain reaction have demon-trated the expression of a range of purinoceptors in de-eloping skeletal muscle, including P2X2, P2X5, P2X6,2Y1, P2Y2 and P2Y4 (Kolb and Wakelam, 1983; Humend Thomas, 1988; Thomas et al., 1991; Henning et al.,992; Meyer et al., 1999a,b; Bo et al., 2000; Ruppelt et al.,001; Ryten et al., 2001; Cheung et al., 2003). Specificoles in skeletal muscle fiber formation and function haveeen identified for two of these receptors. While activationf the P2Y1 receptor has been shown to regulate acetyl-holine receptor (AChR) and acetylcholinesterase (AChE)xpression at NMJs (Choi et al., 2001; Ling et al., 2004),he P2X5 receptor has been implicated in the regulation ofyoblast activity and muscle regeneration (Ryten et al.,002, 2004). The function of the P2X2 receptor is lesslear. However, the singular ability of this receptor type to

nteract with nicotinic AChRs to produce cross-inhibition ofhannel opening (Nakazawa, 1994; Barajas-López et al.,998; Searl et al., 1998; Zhou and Galligan, 1998; Khakht al., 2000), and the timing of P2X2 receptor expression

n development suggests a role for purinergic signalingn the late stages of NMJ formation and patterning.

In the present study we use the P2X2 receptor knock-ut mouse (P2X2

�/�) (Cockayne et al., 2005) to investi-ate the role of the P2X2 receptor in NMJ development andkeletal muscle function.

EXPERIMENTAL PROCEDURES

nimals

2X2�/� mice were generated by introducing a deletion encom-

assing exons 2–11 into the mouse P2X2 gene (see Cockayne etl., 2005 for details).

mmunohistochemistry and histology

our female wild-type and four P2X2�/� mice were killed by CO2

sphyxiation and death was confirmed by cervical dislocation

ccording to Home Office (UK) regulations covering Schedule 1

ved.

pUasMmrlclTsfi(bmpebrUmtN42a�TbSPcwcc

A

Trayamp(ccpuP

M

AwsancacTmpps

An

AaPtw(wS

M

AmwpgTaemgSasPswp

E

Swwgwcqmftu

P

TdmMaasa41esdt

Nt

Ta

M. Ryten et al. / Neuroscience 148 (2007) 700–711 701

rocedures. All experiments conformed to the Royal Free andniversity College Medical School guidelines on the ethical use ofnimals; experiments were designed to minimize the use anduffering of animals. The soleus muscles were rapidly removed.uscle samples used for the assessment of muscle fiber number,uscle fiber size or muscle fiber type were cut such that the mid-belly

egion was isolated, covered with OCT compound and frozen iniquid nitrogen–cooled isopentane. Cryostat sections (12 �m) wereut to produce transverse muscle sections. Sections showing the

argest muscle bulk were collected on gelatinized slides, stained witholuidine Blue and used for assessment of muscle fiber number andize. For assessment of muscle fiber type by demonstration of myo-brillar ATPase we followed the method described by Dubowitz1985) with preincubation of fresh frozen muscle sections in alkaliuffer (pH 9.4). In order to immunostain for fast skeletal muscleyosin, cryostat sections were fixed in 4% formaldehyde in 0.1 Mhosphate buffer for 2 min, rinsed several times in phosphate-buff-red saline (PBS) and incubated overnight with a monoclonal anti-ody (diluted 1:100 in 10% goat serum in PBS) that exclusivelyecognizes fast skeletal muscle myosin (Sigma Chemical Co., Poole,K). Immunoreactivity was visualized by incubation with goat anti-ouse fluorescein-conjugated secondary antibody (Stratech Scien-

ific, Newmarket, UK). Muscle samples used for visualization ofMJs were pinned at proximal and distal ends to sylgard and fixed in% paraformaldehyde in PBS for 45 min, rinsed briefly and sunk in0% sucrose in PBS. Following fixation, muscle fibers were teasedpart using fine forceps and incubated with Texas Red–labeled-bungarotoxin (Invitrogen Ltd., Paisley, UK) diluted 1:1000 in PBS.o immunostain for synaptophysin, teased muscle fibers were incu-ated overnight in rabbit anti-synaptophysin antibodies (Synapticystems, Goettingen, Germany) diluted 1:100 in 10% goat serum inBS. Immunoreactivity was visualized using goat anti-rabbit fluores-ein-conjugated secondary antibody (Stratech Scientific). Imagesere captured using a Leica (Heerbrugg, Switzerland) confocal mi-roscope and the sum pixel function Leica analysis software toollapse z-series stacks taken through entire NMJs.

nalysis of AChR expression at NMJs

eased muscle fibers stained with Texas Red–labeled �-bunga-otoxin and photographed as described above, were used fornalysis of AChR expression at NMJs. Using Scion Image anal-sis software (NIH, USA) the number of AChR clusters per NMJnd total area of AChR expression per NMJ were measured. Ainimum of 45 NMJs (taken from at least three mice) were sam-led to make up each experimental group. End-plate morphologycluster number) in wild-type and P2X2

�/� soleus muscles wereompared using the Mann-Whitney test since the data werelearly not normally distributed. The total area of AChR expressioner NMJ in wild-type and P2X2

�/� soleus muscles was comparedsing a two-tailed unpaired Student’s t-test. A probability level of�0.05 was taken as significant in all tests.

easurement of muscle fiber number

minimum of five samples, each taken from different mice (n�4),as used to assess muscle fiber number in wild-type or P2X2

�/�

oleus muscles. Soleus muscle sections prepared as describedbove were used for analysis and photographed under �2.5 mag-ification using a Nikon digital camera. Muscle fiber numbers wereounted by enlarging the screen image and using a grid (to preventccidental re-counting of fibers). Wild-type and P2X2

�/� soleus mus-le fiber numbers were compared using an unpaired Student’s t-test.he numbers of fast (type II) muscle fibers, as demonstrated byyofibrillar ATPase were measured in a similar manner and ex-ressed as a % of the total number of muscle fibers counted. Theercentage of fast type muscle fibers in wild-type and P2X �/�
2oleus muscle was compared using an unpaired Student’s t-test. e
ssessment of the incidence of centrallyucleated fibers

minimum of four samples from different mice (n�4) was used tossess the incidence of centrally nucleated fibers in wild-type or2X2

�/� soleus muscles. Using the method described above, theotal number and the number of centrally nucleated muscle fibersere counted. The incidence of centrally nucleated muscle fibers

expressed as a percentage of the total number of muscle fibers) inild-type and P2X2

�/� muscle was compared using an unpairedtudent’s t-test.

easurement of muscle fiber cross-sectional area

minimum of five different samples, each taken from differentice, was used to assess muscle fiber cross-sectional area inild-type or P2X2

�/� soleus muscles. Soleus muscle sectionsrepared as described above were used for analysis, and photo-raphed under �20 magnification using a Nikon digital camera.wo sections per muscle were analyzed. Images were viewed oncomputer screen and a 200 �m�200 �m grid was applied to

ach image. Random selection of muscle fibers was achieved byaking fiber area measurements only on fibers overlying points ofrid intersections. Cross-sectional fiber area was measured usingcion Image analysis software and expressed in �m2. In this wayt least 70 muscle fibers (approximately 8% of muscle fibers) wereampled from each image and a total of 702 wild-type and 7192X2

�/� soleus muscle fibers were measured. Average cross-ectional fiber areas (calculated for each muscle sample) fromild-type (n�5) and P2X2

�/� (n�5) soleus muscles were com-ared using a two-tailed unpaired Student’s t-test.

lectron microscopy

oleus muscles from three wild-type and three P2X2�/� animals

ere fixed in 2% glutaraldehyde in 0.1 M cacodylate buffer,ashed, post-fixed in 1% osmium tetroxide (OsO4), dehydrated inraded alcohol and embedded in resin. Thin longitudinal sectionsere cut at 80 nm thick, stained with uranyl acetate and leaditrate and examined under a JEM-1010 electron microscope. Touantify fold density, the total length of terminal contact waseasured on photomicrographs and the number of junctional

olds that could be seen was recorded. Measurements from wild-ype and P2X2

�/� muscles were compared using a two-tailednpaired Student’s t-test.

hysiology

he soleus muscle from five wild-type and five P2X2�/� mice was

issected free and placed in continually gassed (95% O2/5% CO2)odified Krebs solution (mM): NaCl, 133; KCl, 4.7; NaHCO3, 16.4;gSO4, 0.6; NaH2PO4, 1.4; glucose, 7.7; and CaCl2, 2.5; pH 7.3t 37�1 °C. One end of the muscle was attached to a rigid supportnd the other to a Grass FT03C force displacement transducer viailk ligatures and mounted in 10 ml organ baths. Mechanicalctivity was recorded using PowerLab Chart for Windows (Version; ADInstruments New South Wales, Australia). An initial load ofg was applied to each soleus muscle, which was allowed to

quilibrate for 60 min. Contractions were elicited by electricaltimulation of the muscle via two platinum wire rings 2.5 mm iniameter and 1 cm apart through which the soleus muscles werehreaded.

erve-mediated and muscle-mediatedwitch responses

witch contractile responses due to the stimulation of nerves werechieved by using: 100 V, 0.1 ms duration and one twitch/min. By

xtending the duration of stimulation to 5 ms direct stimulation of

mpcMtdct

T

TmPifcc(

Se

Tre2tsbutcuDpFgs(gcrrp

PW(WPt(si

Ps

Uomw

mpna

FslrTrws(wab

M. Ryten et al. / Neuroscience 148 (2007) 700–711702

uscle was achieved. In each case the means of five responseser soleus muscle were taken and a mean contraction (in g) wasalculated for each muscle (from a total of five animals per group).uscle-mediated tetanic contractions were measured in both wild-

ype and P2X2�/� soleus muscles by stimulating with 100 V, 5 ms

uration at 50 Hz, until a maximum contraction was achieved. Allontractions were compared using a two-tailed unpaired Student’s-test.

ubocurarine sensitivity

he inhibitory effect of tubocurarine (0.1–3 mM) against nerve-ediated twitch responses was compared in wild-type and2X2

�/� soleus muscles. The effect of tubocurarine on muscle-nduced twitches was calculated as mean % reduction oforce�S.E.M. (n animals) and a concentration–response curveonstructed for wild-type and P2X2

�/� soleus muscles. Theurves were compared using a two-way analysis of varianceANOVA) followed by a Bonferroni’s post hoc test.

ingle channel recording from muscle fibernd-plates

he foot muscles (lumbrical and flexor digitorum brevis) wereemoved from at least six animals per group and incubated inxtracellular solution containing collagenase (Worthington CLS2,mg ml�1) for 1 h at 37 °C. The muscles were then rinsed

horoughly in enzyme free solution and triturated gently to releaseingle muscle fibers. The fibers were allowed to adhere to theottom of 35 mm diameter culture dishes (Falcon), and viewedsing an inverted microscope with a 40� phase contrast objec-ive. Single channel recording in the cell attached or inside outonfiguration was carried out at room temperature (21–23 °C),sing an Axopatch 1B (Axon Instruments, Union City, CA, USA).ata were filtered at 5 kHz, digitized at 10 kHz and stored on aersonal computer using pClamp v8, for offline analysis usingetchan v6. Electrodes were pulled from thick wall borosilicatelass and had a resistance of 8–12 M� when filled with bathingolution containing 1 �M ACh. The bathing solution consisted ofmM): NaCl 154; KCl 4.7; MgCl2 1.2; CaCl2 2.5; HEPES 10 andlucose 5.6; the pH was adjusted to 7.4 using NaOH. The slopeonductance of the single channels was determined from linearegression of single channel current amplitude, recorded duringamp changes in membrane potential, versus patch membraneotential.

RESULTS

2X2�/� mice were fertile with no signs of gross pathology.

e did not detect any significant differences in body mass27.8 g�1.7 g, wild-type; 27.6 g�0.5 g, P2X2

�/� mice).hile differences in the muscle masses of wild-type and2X2

�/� extensor digitorum longus (10.9�0.7 mg, wild-ype; 9.68�0.5 mg, P2X2

�/�) and soleus muscles9.1�0.44 mg, wild-type; 8.2�0.32 mg, P2X2

�/�) wereuggestive of muscle atrophy, values did not reach signif-cance (P�0.15 for soleus muscle).

2X2 receptor expression in developingkeletal muscle

sing immunohistochemistry we confirmed the expressionf P2X2 receptor protein in developing wild-type skeletaluscle (Fig. 1a). Immunostaining for the P2X receptor
2as not observed in embryonic day 14 (E14) skeletal w
uscle, this receptor was strongly expressed in wild-typeostnatal day 3 (P3), P7 and P15 muscle (Fig. 1a). Immu-ostaining was localized to the muscle fiber membranesnd spread along the length of the muscle fibers. There

ig. 1. P2X2 receptor expression in developing P2X2�/� and wild-type

keletal muscle. (a) Immunostaining of wild-type control E16 lowerimb muscles, P7, P15 and P21 soleus muscle. Expression of the P2X2eceptor was found on wild-type skeletal muscle at P7 and P15 only.his expression was localized to muscle fiber membranes with noeceptor expression in between muscle fibers at any stage. (b) Thereas no evidence of P2X2 receptor expression in skeletal muscle at anytage of development (E16, P7, P15 and P21). Scale bars�50 �m.c) Double-staining for AChRs (red) and P2X2 receptors (green) inild-type soleus muscle at P15. (d) Double-staining for AChRs (red)nd P2X2 receptors (green) in wild-type soleus muscle at P15. Scalears�50 �m. Wild-type, (�/�); P2X2

�/�, (�/�).

as no evidence of specific localization of P2X2 receptor

es1oorbr(

Am

IdpaA

pPoatPAPmwitemQinpEs(dcePwrt(

wTttAiPAaol

e(ws

Ai

IaSipANd3p(tPplsdtip

CPm

T�emgwpmc(vi4cwpPc(

MP

Ert


xpression at NMJs as shown by double-staining P15oleus muscle for both the P2X2 receptor and AChRs (Fig.c). P2X2 receptor expression was not maintained in P21r in adult wild-type skeletal muscle (Fig. 1a). This patternf P2X2 receptor expression is not consistent with specificeceptor expression at NMJs or in motor nerve bundlesetween muscle fibers. There was no evidence of P2X2eceptor expression at any age in P2X2

�/� skeletal muscleFig. 1b and d).

bnormalities in NMJ structure in P2X2�/� skeletal

uscle visualized by light microscopy

n wild-type soleus muscle we observed the previouslyescribed transformation of endplates from simple ovallaques of AChRs at high density, to perforated plaques,nd ultimately to continuous pretzel-shaped areas ofChR expression (Fig. 2a).

There appeared to be no differences in endplate mor-hology between wild-type and P2X2

�/� skeletal muscle at7. However, P2X2

�/� endplates examined at later stagesf development (P15 and adult) were less organized thange-matched controls. In particular, the smooth outline ofhe endplate regions observed in wild-type muscle fibers at15 (representing the demarcation between high and lowChR expression) appeared blurred in endplates from2X2

�/� muscle fibers (Fig. 2a). Abnormalities in NMJorphology in P2X2

�/� muscle became more prominentith increasing age (P15 and adult). Unlike the endplates

n adult wild-type soleus muscle, which consisted of con-inuous pretzel-shaped areas of AChR expression, manyndplates in adult P2X2

�/� soleus muscle were frag-ented with multiple separate clusters of AChRs (Fig. 2a).uantitative analysis of endplate structure confirmed this

mpression, demonstrating a significant increase in theumber of AChR clusters/endplate in P2X2

�/� as com-ared with wild-type soleus muscle (P�0.00081) (Fig. 2b).ndplates in P2X2

�/� muscle (54.9�2.0 �m) were alsoignificantly longer than those in age-matched wild-type49.3�1.6 �m) soleus muscle (P�0.031). Despite theseifferences in endplate morphology, there was no signifi-ant difference in the total area of AChR expression perndplate in wild-type (527�27 �m2) as compared with2X2

�/� (561�23 �m2) soleus muscle (Fig. 2c). Similarlye found no significant difference in the intensity of fluo-

escence (visualized by Texas Red–labeled �-bungaro-oxin binding) at wild-type (173.2�1.8 units) and P2X2

�/�

176.9�4.4 units) NMJs.Adult skeletal muscle fibers were also double-stained

ith anti-synaptophysin to visualize nerve terminals andexas Red–labeled bungarotoxin to visualize post-synap-

ic AChRs. In wild-type soleus muscle NMJs were charac-erized by apposition of nerve terminal branches andChR-rich segments of post-synaptic membrane, visual-

zed as yellow staining (Fig. 2d). However, in some2X2

�/� NMJs there was misalignment between theChR-rich regions of the post-synaptic membrane (red)nd nerve terminal branches (green) (Fig. 2d). Analysisf en face images of NMJs demonstrated significantly

ess (P�0.013) alignment of synaptophysin and AChR P

xpression (as visualized by yellow staining) in P2X2�/�

68.8�3.2% of total AChR expression) as comparedith wild-type NMJs (81.0�3.1% of total AChR expres-ion).

bnormalities in the ultrastructure of NMJsn P2X2

�/� soleus muscle

n wild-type soleus muscles, NMJs (at all ages) were char-cterized by the presence of nerve terminals capped bychwann cell processes and post-synaptic muscle special-

zation, including junctional folds, a raised myofilament-oor sole plate and aggregates of myonuclei (Fig. 3).lthough junctional folds were also observed in P2X2

�/�

MJs examined at P15, there was a reduction in theensity and depth of folds in the P2X2

�/� muscles (Fig.a). This difference in fold accumulation or maintenanceersisted and became more prominent in adult muscleFig. 3b). There were 5.2�0.2 junctional folds/�m in wild-ype adult NMJs, but only 3.4�0.2 junctional folds/�m in2X2

�/� NMJs (Fig. 3c). In areas of infolding of theostjunctional membrane, the folds appeared to be shal-

ower than those in wild-type NMJs (Fig. 3a and b). Mosturprising was the apparent variability in post-junctionalifferentiation within some NMJs. Whereas the majority ofhe post-junctional membrane apposing the nerve terminaln Fig. 3b appears flat (with no junctional folds), a smallortion is extensively folded.

hannel properties of AChRs expressed in2X2 receptor positive wild-type and P2X2

�/�

uscle fibers

he properties of the specific AChR subtypes (embryonic,subunit-containing or adult, � subunit-containing AChRs)xpressed at the endplate are known to influence NMJaturation (Missias et al., 1997). Therefore we investi-ated the channel properties of the AChRs expressed atild-type and P2X2

�/� skeletal muscle. Single channelroperties of digitorum brevis end-plates in P5 wild-typeuscle fibers had the characteristics of typical embryonic

hannels, with longer openings and a lower conductanceFig. 4a, Table 1). In contrast, adult wild-type mice re-ealed AChR channels with typical adult characteristics,

ncluding short opening times and a high conductance (Fig.a, Table 1). The mean open time and conductance forhannels from adult digitorum brevis P2X2

�/� endplatesere similar to those from adult wild-type muscle. Theresence of adult-type AChRs in both wild-type and2X2

�/� tissue was confirmed by the absence of signifi-ant immunostaining for the � (embryonic) AChR subunitFig. 4b).

uscle fiber atrophy in soleus muscle from2X2

�/� mice

xamination of P2X2�/� soleus muscle sections did not

eveal evidence of muscle necrosis (excessive connectiveissue) or regeneration (central nuclei) at any age (E14,

4, P7, P14, P21 and adult). This was confirmed by mea-

FRPegsd((at


ig. 2. Abnormalities in NMJ structure in P2X2�/� skeletal muscle visualized by confocal microscopy. (a) Representative endplates stained with Texas

ed–labeled �-bungarotoxin for AChRs from P7, P15 and adult wild-type and P2X2�/� soleus muscles. Whereas at P7 endplates in control and

2X2�/� skeletal muscle were indistinguishable, by P15 some endplates in P2X2

�/� skeletal muscle appeared ragged. Furthermore, while control adultndplates consisted of a single AChR cluster per endplate (white arrowheads), in P2X2

�/� muscle the number of clusters per endplate was significantlyreater. Scale bars�20 �m. (b) Box and whisker diagram to show the number of clusters of AChR expression per adult endplate. (c) Scatter plot tohow the maximum distance between areas of AChR expression within a single adult endplate in wild-type and P2X2

�/� soleus muscle. No significantifference in the total area of AChR expression in wild-type and P2X2

�/� was found, suggesting that the total number of AChRs expressed was similar.d) Representative NMJs from adult wild-type and P2X2

�/� soleus muscles double-stained for the nerve terminal protein, synaptophysin (green) and AChRsred). Precise apposition of nerve terminals and AChR-rich regions of the post-synaptic membrane (as visualized by yellow staining) was a feature of controldult NMJs. However, in P2X �/� muscle there were areas of AChR expression on muscle without overlying nerve fibers (red staining only) and nerve
2
erminals that were not apposed by AChR-rich muscle regions (green staining only). Scale bars�20 �m. Wild-type (�/�), P2X2�/� (�/�).

FPpwetP


ig. 3. Abnormalities in the ultrastructure of NMJs in P2X2 receptor-deficient soleus muscle. (a) Electron micrographs of representative NMJs in wild-type and2X2

�/� soleus muscles at P15 demonstrating the development of deep postsynaptic folds. Whereas wild-type NMJs developed deep post-synaptic folds,ost-synaptic folds in P2X2

�/� soleus muscle at P15 were shallow and fewer in number. Scale bar�1 �m. (b) Electron micrographs of representative NMJs in adultild-type and P2X2

�/� soleus muscle. At wild-type NMJs the post-synaptic membrane of the muscle fiber (M) underlying the motor nerve terminal (N) wasxtensively folded. NMJs in P2X2

�/� soleus muscle were characterized by reduced folding of the post-synaptic membrane of the muscle (M) underlying nerveerminals (N). Scale bar�1 �m. (c) Scatter plot to show the density of post-synaptic folds at NMJs in wild-type and P2X �/� adult soleus muscle. Wild-type (�/�),
22X2
�/� (�/�).

samasHa5tmtc(mnfiP

Ci

Wat5dbmsvbsi5itt(

tttcnP(o(smm1

Fs

Iwd(KmPfipapoe(

T(

M

AA

FmaesRb�tsdpb


uring the incidence of centrally nucleated muscle fibers indult (2–2.5 month old) wild-type and P2X2

�/� soleususcle. The incidence of central nuclei in both wild-typend P2X2

�/� muscle was low (0.40�0.14% in wild-typeoleus muscle; 0.42�0.06% in P2X2

�/� soleus muscle).owever, muscle fibers in adult P2X2

�/� soleus muscleppeared smaller than wild-type soleus muscle fibers (Fig.a). Measurement of cross-sectional muscle fiber areas inhe mid-belly region of adult wild-type and P2X2

�/� soleususcles confirmed this finding. Muscle fiber cross-sec-

ional areas in adult P2X2�/� (630�16 �m2) were signifi-

antly smaller (P�0.038) than those in wild-type731�37 �m2) soleus muscle, indicating a reduced rate ofuscle fiber growth in P2X2

�/� soleus muscle. There waso significant difference in the total number of musclebers present in wild-type (859�45) compared with

ig. 4. Properties of AChRs expressed in wild-type and P2X2�/�

uscle fibers could be detected. (a) Single channel recordings fromdult wild-type, adult P2X2

�/� and P5 mouse digitorum brevis musclendplates in the presence of 1 �M ACh. (b) Embryonic-type � AChRubunits were not expressed in wild-type or P2X2

�/� soleus muscle.epresentative NMJs from adult wild-type and P2X2

�/� digitorumrevis muscle double-stained for AChRs with Texas Red–labeled-bungarotoxin (red) and a specific antibody against the � subunit of

he AChR (green). The absence of green and yellow staining demon-trates that the AChRs expressed in both wild-type and P2X2

�/�

igitorum brevis muscle did not contain the � subunit, suggesting theresence of mature as opposed to embryonic-type AChRs. Scalears�25 �m. Wild-type (�/�), P2X2

�/� (�/�).

2X2�/� (893�36) soleus muscle (Table 2).

D

hanges in response to tetanic stimulationn P2X2


e assessed muscle function by comparing nerve-medi-ted twitch, muscle-mediated twitch and tetanic contrac-ions in adult wild-type and P2X2

�/� soleus muscles (Fig.b, Table 2). We found no difference between the forceseveloped by stimulating the muscle directly or indirectlyy stimulation of the nerve, suggesting the absence ofuscle denervation in either wild-type or P2X2

�/� adultoleus muscle. We assessed NMJ function further by in-estigating sensitivity to the AChR-specific antagonist, tu-ocurarine. While we were unable to demonstrate anyignificant difference between the IC50 values measuredn P2X2

�/� (IC50�5.76�0.14 �M) and wild-type (IC50�.89�0.08 �M) soleus muscle, the results did suggest an

ncrease in AChR reserve at P2X2�/� NMJs (as shown by

he increase in the concentration of tubocurarine requiredo inhibit muscle contraction on indirect nerve stimulation)Fig. 5b).

No statistically significant differences could be de-ected between the force developed by single twitch con-ractions (generated by direct muscle stimulation) in wild-ype and P2X2

�/� soleus muscles (Table 2). However,omparison of twitch contraction times demonstrated sig-ificantly (P�0.0001) shorter times to peak contraction in2X2

�/� (86.69�1.53 ms) as compared with wild-type97.38�1.57 ms) soleus muscle suggesting the presencef a higher number of fast type fibers in P2X2

�/� muscleFig. 5c). Furthermore, the force developed during tetanictimulation was significantly (P�0.043) reduced in soleususcle from adult P2X2

�/� (6.8�0.8 g, 584�10 g/cm2 ofuscle) as compared with wild-type control (9.7�1.4 g,115�135 g/cm2 of muscle) soleus muscle (Table 2).

ast (type II) muscle fibers in wild-type and P2X2�/�

oleus muscle

mmunostaining of wild-type and P2X2�/� soleus muscle

ith antibodies specific to fast skeletal muscle myosinemonstrated the presence of fast (green) and slowblack) fibers (Fig. 6a). Consistent with the findings ofugelberg (1976), both wild-type and P2X2

�/� soleususcle consisted predominantly of type II fibers (green) at7, but showed an increase in the proportion of type Ibers (black) with age (Fig. 6a). However, by P14 theroportion of type II fibers (determined by immunostaining)ppeared to be greater in P2X2

�/� soleus muscle com-ared with age-matched wild-type controls (Fig. 6a). Dem-nstration of myofibrillar ATPase gave a similar pattern ofxpression with an increase in the expression of fast fibersblack) relative to slow fibers (white) (Fig. 6b). Quantitative

able 1. Channel properties of AChRs expressed in wild-typeP2X2

�/�) and P2X2�/� muscle fibers

uscle Mean open times (ms) Conductance (pS)

dult P2X2�/� (n�7) 0.46�0.05 45.6�2.0

dult P2X2�/� (n�7) 0.63�0.05 40.2�3.0�/�

ay 5 P2X2 (n�6) 4.80�1.20 36.6�1.6

atotta6mm

Atmttss

FsmPa emonstram

T

TFTCTTC


nalysis of muscle fiber type in adult muscle demonstratedhe presence of a significantly higher (P�0.016) proportionf fast muscle fibers in adult P2X2

�/� compared with wild-ype soleus muscle. Whereas 57.0�2.2% of adult wild-ype soleus muscle fibers were type II, 71.8�4.0% ofdult P2X2

�/� soleus muscle fibers were fast fibers (Fig.c). Thus, the developmental transition from fast to slowuscle fibers failed to fully occur in P2X2

�/� soleususcle.

ig. 5. Muscle fiber atrophy and changes in response to tetanic stimoleus muscle stained with Toluidine Blue. While there was no evidenuscle, there was evidence of muscle fiber atrophy. Scale bars�1002X2

�/� (open circles) adult soleus muscle on indirect nerve stimulatind P2X2

�/� (gray line) soleus muscle on direct muscle stimulation, duscle. Wild-type (�/�), P2X2

�/� (�/�).

able 2. Structural and contractile properties of wild-type (P2X2�/�) a

Adult P2X2�/�

otal no. of fibers 859�45 (n�4iber cross-sectional area (�2) 731�37 (n�4witch tension (g) 0.82�0.11 (n�orrected twitch tension (g/cm2) 92.5�11.4 (n�ime to peak contraction (ms) 97.38�1.57 (n�etanic tension (g) 9.7�1.4 (n�

2
orrected tetanic tension (g/cm ) 1115�135 (n�4)
DISCUSSION

lthough ATP is well recognized as an important neuro-ransmitter in the CNS and peripheral nervous systemediating fast synaptic signaling (see Burnstock, 2007),

he role of purinergic signaling in the formation and main-enance of synapses has not been fully investigated. In thistudy P2X2 receptor-deficient mice were used to demon-trate a role for ATP signaling in the formation of mature

soleus muscle from P2X2�/� mice. (a) Wild-type and P2X2

�/� adultcle necrosis or regeneration of muscle fibers in adult P2X2

�/� soleusDose inhibition curve for tubocurarine in wild-type (filled circles) andpresentative examples of twitch contractions in wild-type (black line)

ting a significantly shorter time to peak contraction in P2X2�/� soleus


Adult P2X2�/� P-value

893�36 (n�4) NS630�16 (n�4) P�0.05

0.72�0.05 (n�12) NS70.4�1.0 (n�4) NS

86.69�1.53 (n�12) P�0.00016.8�0.8 (n�12) P�0.05

ulation ince of mus

�m. (b)on. (c) Re

nd P2X2

))10)4)10)

10)

584�10 (n�4) P�0.05

NP

i

FPa(ha ld-type (�


MJs. Our results show that the normal expression of the

ig. 6. Changes in muscle fiber type in P2X2�/� soleus muscle. (a)

2X2�/� P7, P15 and adult soleus muscle. Scale bars�50 �m. (b) Sta

dult wild-type and P2X2�/� soleus muscle to demonstrate the prese

c) Histogram to show the numbers of fast (type II) muscle fibersistochemistry) expressed as a % of the total number of muscle fiberss compared with wild-type adult soleus muscle is denoted by (*). Wi

2X2 receptor protein on developing skeletal muscle fibers N

s required for the formation and maintenance of mature

taining for fast type skeletal muscle myosin (green) in wild-type andPase enzyme histochemical staining (at pH 9.4) of whole sections ofpe I (white) and type II (black) muscle fibers. Scale bars�500 �m.wild-type and P2X2

�/� soleus muscle (as determined by enzymenificant increase (P�0.05) in the proportion of fast fibers in P2X2

�/�

/�), P2X2�/� (�/�).

Immunosndard ATnce of tyin adult. The sig

MJs and that absence of this receptor results in disorga-

nrgc

s2dnfiffisdotS(tpctai

wseermnmdpfcmcasgm

dbtedbn

pdmfqto

am

mcscnfamtgvtrtmsimtiPtaiecmm(safsatsl

apnOggHwoNi2ssdet


ized synapses. These defects in NMJ formation mayesult in less reliable neuromuscular transmission, as sug-ested by the incomplete transition from fast to slow mus-le fiber type and decreased muscle fiber size.

Consistent with previous studies showing the expres-ion of P2X2, P2X5 and P2X6 receptors (Ryten et al., 2001,004), we demonstrated expression of P2X2 receptors oneveloping wild-type mouse skeletal muscle using immu-ohistochemistry. This receptor was expressed on muscleber membranes in early postnatal skeletal muscle. Weound no evidence of P2X2 receptor expression in nervebers or surrounding Schwann cells, nor has such expres-ion been reported in mammalian NMJs in the literature toate. However, functional P2X receptors have been dem-nstrated in peri-synaptic Schwann cells in frog NMJs, sohe possibility that these receptors are present onchwann cells in mammalian NMJs cannot be excluded

Robitaille, 1995). Nonetheless, these findings suggesthat while the P2X2 receptor was not specifically ex-ressed at mature muscle endplates, the observedhanges in NMJ structure and skeletal muscle function inhe P2X2

�/� mouse were most likely to be due to thebsence of P2X2 receptor-mediated signaling on develop-

ng muscle cells.The timing of P2X2 receptor expression is consistent

ith the pattern of abnormalities observed in P2X2�/�

keletal muscle. The P2X2 receptor was not expressed inmbryonic and adult skeletal muscle, but was present inarly postnatal development. Therefore absence of thiseceptor would be expected to affect the later stages ofuscle development. In keeping with this view, we foundo change in muscle fiber number in P2X2

�/� soleususcle, the total number of muscle fibers formed beingependent on myotube formation, a process which is com-leted by birth. Similarly, AChR clustering which rapidlyollows myotube formation in the embryo (Sanes and Li-htman, 2001), occurred normally in P2X2

�/� skeletaluscle. However, abnormalities in NMJ maturation (in-

luding the formation of a complex pretzel-like shape fromsimple oval cluster of AChRs and the formation of post-

ynaptic folds), muscle fiber typing and muscle fiberrowth were all affected in skeletal muscle of P2X2

�/�

ice.Abnormalities in P2X2

�/� skeletal muscle, includingisorganized NMJs and an increase in fast muscle fibers,ecame apparent from P15. Staining for AChRs revealedhe ragged appearance of many of the endplates andlectron microscopy demonstrated reduced numbers andepth of post-synaptic folds. However, these differencesetween wild-type and P2X2

�/� NMJs became more sig-ificant in adult tissues.

This might suggest a role for P2X2 receptor-mediatedurinergic signaling in NMJ maintenance. While we did notetect P2X2 receptor expression in healthy adult skeletaluscle, this observation could be explained by the need

or re-expression of this receptor on muscle injury (a fre-uent occurrence in the lifetime of animals). This interpre-ation of the findings is supported by recent studies dem-

nstrating re-expression of the P2X2 receptor in regener- s

ting skeletal muscle fibers in the mdx (mouse model) ofuscular dystrophy (Ryten et al., 2004).

The most striking abnormalities in P2X2�/� skeletal

uscle were observed at NMJs. Adult P2X2�/� NMJs

onsisted of significantly more AChR clusters/endplatepread over a wider area as compared with wild-typeontrols. There was mis-apposition between pre-synapticerve terminal specialization and post-synaptic muscle dif-erentiation in P2X2

�/� NMJs. Most significantly there wasmarked reduction in post-synaptic folding at P2X2

�/�

uscle endplates. The absence of any physiological datao suggest partial or complete muscle denervation sug-ests that the abnormalities in NMJ structure did not pre-ent neurotransmission. However, the reliability of neuro-ransmission may have been affected. While effective neu-otransmission occurs (in embryonic life) without many ofhe pre- and post-synaptic specializations associated withature NMJs, these specializations are required to ensure

ynapse reliability (Slater, 2003). Therefore the abnormal-ties in NMJ structure present in adult P2X2

�/� skeletaluscle might still be expected to result in abnormal func-

ion. Indeed the generalized atrophy of muscle fibers andncreased percentage of fast type muscle fibers in adult2X2

�/� soleus muscle is not only consistent, but sugges-ive of abnormal neuromuscular transmission. Whereasdult wild-type and P2X2

�/� soleus muscle contained sim-lar numbers of fibers, P2X2 receptor-deficient muscle gen-rated lower forces on tetanic stimulation. This differenceould be explained by significant fiber atrophy in P2X2

�/�

uscle secondary to reduced activity. Reliable neurotrans-ission is also important in directing muscle fiber typing

Carrasco and English, 2003). The transition from fast tolow type muscle fibers is induced by increased musclectivation during development (Kugelberg, 1976). There-ore, the failure of complete fiber type transition (fast tolow) in P2X2

�/� soleus muscle strongly suggests that thebnormal NMJs of P2X2

�/� mice are unable to effectivelyransmit this increased activity. Thus, the defects in NMJtructure observed at EM and by confocal microscopy are

ikely to be of functional significance.The abnormalities in NMJ structure, muscle fiber type

nd size demonstrated in P2X2�/� muscle could be ex-

lained by a single role for P2X2 receptor-dependent sig-aling in the development and maintenance of NMJs.ther explanations might include the possibility of a moreeneral role for P2X2 receptors in muscle activation, sug-esting that changes in NMJ structure are secondary.owever, since the most pronounced changes in muscleere observed at the NMJs, we feel that the primary effectf P2X2 receptor-dependent signaling is likely to be onMJ morphology. Unlike P2Y1 receptor activation, which

ncreases the expression of AChR protein (Choi et al.,001; Ling et al., 2004), we were unable to demonstrate aignificant change in the total expression of AChRs onoleus muscle in P2X2

�/� mice. We found no significantifference in the total area or density of AChR staining atndplates, and although there was a mild reduction inubocurarine sensitivity in P2X �/� muscle, we found no

2ignificant differences in the IC50 values in wild-type and

Pdanlbpptv

pmotaotcwsAiascvstCssms

IrierctbnlssHmtccpsw

B

B

B

B

C

C

C

C

D

H

H

H

H

J

K

K

K

L

M

M

M


2X2�/� soleus muscle. Nor did we detect any significant

ifferences in the AChR subtypes expressed at wild-types compared with P2X2

�/� endplates (using single chan-el recording or immunohistochemistry). Although these

atter experiments were conducted on the flexor digitorumrevis muscles, where NMJ structure was not investigated,revious studies suggest that P2X2 receptor expression isart of the developmental program of all muscle fibers andhat extrapolating findings to the soleus muscle may bealid (Ryten et al., 2001, 2004).

Thus, P2X2 receptor-dependent signaling appears tolay a specific role in directing some features of endplateorphology. This function is consistent with the well-rec-gnized importance of innervation in NMJ formation, andhe fact that while the P2X2 receptor may be expressedcross the muscle membrane, the most significant releasef ATP is likely to be at the motor nerve terminal where thisransmitter is co-released with ACh. Furthermore, the spe-ific properties of the P2X2 receptor make it particularlyell-suited to a role in directing AChR distribution. Theingular ability of this receptor to interact with nicotinicChRs in a density-dependent manner to produce cross-

nhibition of channel opening (Nakazawa, 1994; Searl etl., 1998; Zhou and Galligan, 1998; Khakh et al., 2000)uggests a mechanism by which P2X2 receptor activationould mediate localized effects on AChR expression. Pre-ious studies conducted by O’Malley et al. (1997), demon-trating that ATP can stabilize AChRs on cultured myo-ubes to increase receptor half-life, support this theory.ertainly a role for purinergic signaling in ensuring AChRtability could explain the timing of P2X2 receptor expres-ion in early postnatal life, the need for re-expression onuscle regeneration and the variability in NMJ structure

een in the P2X2�/� mice.

CONCLUSION

n summary, we demonstrate that absence of the P2X2eceptor on skeletal muscle results in significant abnormal-ties in muscle structure and function. These effects can bexplained most simply by proposing a role for the P2X2eceptor in the normal development of the NMJ. It is be-oming increasingly clear that the development and main-enance of synapses is a complex process requiring aalance between stabilizing and de-stabilizing factors. Aumber of other factors have been shown to influence the

atter steps of NMJ formation, including laminin �4 expres-ion (Patton et al., 2001) and cholinergic neurotransmis-ion (Misgeld et al., 2002, 2005; Brandon et al., 2003).owever, much less is known about the maturation andaintenance of synapses than about the initial steps in

heir formation (Sanes and Lichtman, 1999). Thus, theoordinated action of a number of signaling systems, in-luding the stabilizing effects of P2X2 receptor-mediatedurinergic signaling, is likely to be necessary to ensureynapse plasticity not only at the NMJ, but also in the CNS

here the P2X2 receptor is also strongly expressed.

REFERENCES

arajas-López C, Espinosa-Luna R, Zhu Y (1998) Functional interac-tions between nicotinic and P2X channels in short-term cultures ofguinea pig submucosal neurons. J Physiol (Lond) 513:671–683.

o X, Schoepfer R, Burnstock G (2000) Molecular cloning and char-acterization of anovel ATP P2X receptor subtype from embryonicchick skeletal muscle. J Biol Chem 275:14401–14407.

randon EP, Lin W, D’Amour KA, Pizzo DP, Dominguez B, Sugiura Y,Thode S, Ko CP, Thal LJ, Gage FH, Lee KF (2003) Aberrantpatterning of neuromuscular synapses in choline acetyltrans-ferase-deficient mice. J Neurosci 23:539–549.

urnstock G (2007) Physiology and pathophysiology of purinergicneurotransmission. Physiol Rev 87:659–797.

arrasco DI, English AW (2003) Neurotrophin 4/5 is required for thenormal development of the slow muscle fibre phenotype in the ratsoleus. J Exp Biol 206:2191–2200.

heung KK, Ryten M, Burnstock G (2003) Abundant and dynamicexpression of G protein-coupled P2Y receptors in mammaliandevelopment. Dev Dyn 228:254–266.

hoi RC, Man ML, Ling KK, Ip NY, Simon J, Barnard EA, Tsim KW(2001) Expression of the P2Y1 nucleotide receptor in chick muscle:its functional role in the regulation of acetylcholinesterase andacetylcholine receptor. J Neurosci 21:9224–9234.

ockayne DA, Dunn PM, Zhong Y, Hamilton SG, Cain GR, Knight GE,Ruan H-Z, Ping Y, Nunn P, Bei M, McMahon SB, Burnstock G,Ford APDW (2005) P2X2 knockout mice and P2X2/P2X3 doubleknockout mice reveal a role for the P2X2 receptor subunit inmediating multiple sensory effects of ATP. J Physiol 567:621–639.

ubowitz V (1985) Muscle biopsy: a practical approach, 2nd ed, pp19–40. London: Bailliere Tindall.

eilbronn E, Eriksson H (1998) Second messengers mobilized by ATPand ACh in the myotube/muscle fiber. In: Neuromuscular junction(Sellin LC, Libelius R, Thesleff S, eds), pp 395–404. Amsterdam,The Netherlands: Elsevier Science Publishers.

enning RH (1997) Purinoceptors in neuromuscular transmission.Pharmacol Ther 74:115–128.

enning RH, Nelemans A, van den Akker J, den Hertog A (1992) Thenucleotide receptors on mouse C2C12 myotubes. Br J Pharmacol106:853–858.

ume RI, Thomas SA (1988) Multiple actions of adenosine 5=-triphos-phate on chick skeletal muscle. J Physiol 406:503–524.

iang T, Yeung D, Lien CF, Gorecki DC (2005) Localized expressionof specific P2X receptors in dystrophin-deficient DMD and mdxmuscle. Neuromuscul Disord 15:225–236.

hakh BS, Zhou X, Sydes J, Galligan JJ, Lester HA (2000) State-dependent cross-inhibition between transmitter-gated cation chan-nels. Nature 406:405–410.

olb HA, Wakelam MJ (1983) Transmitter-like action of ATP onpatched membranes of cultured myoblasts and myotubes. Nature303:621–623.

ugelberg E (1976) Adaptive transformation of rat soleus motor unitsduring growth. J Neurol Sci 27:269–289.

ing KK, Siow NL, Choi RC, Ting AK, Kong LW, Tsim KW (2004) ATPpotentiates agrin-induced AChR aggregation in culturedmyotubes: activation of RhoA in P2Y1 nucleotide receptor sig-naling at vertebrate neuromuscular junctions. J Biol Chem279:31081–31088.

eyer MP, Clarke JD, Patel K, Townsend-Nicholson A, Burnstock G(1999a) Selective expression of purinoceptor cP2Y1 suggests arole for nucleotide signaling in development of the chick embryo.Dev Dyn 214:152–158.

eyer MP, Gröschel-Stewart U, Robson T, Burnstock G (1999b)Expression of two ATP-gated ion channels, P2X5 and P2X6, indeveloping chick skeletal muscle. Dev Dyn 216:442–449.

isgeld T, Kummer TT, Lichtman JW, Sanes JR (2005) Agrin

promotes synaptic differentiation by counteracting an inhibitory

M

M

N

O

P

R

R

R

R

R

R

R

S

S

S

S

S

T

W

Z


effect of neurotransmitter. Proc Natl Acad Sci U S A 102:11088 –11093.

isgeld T, Burgess RW, Lewis RM, Cunningham JM, Lichtman JW,Sanes JR (2002) Roles of neurotransmitter in synapse formation:development of neuromuscular junctions lacking choline acetyl-transferase. Neuron 14:635–648.

issias AC, Mudd J, Cunningham JM, Steinbach JH, Merlie JP, SanesJR (1997) Deficient development and maintenance of postsynapticspecializations in mutant mice lacking an ‘adult’ acetylcholine re-ceptor subunit. Development 124:5075–5086.

akazawa K (1994) ATP activated current and its interaction withacetylcholine activated current in rat sympathetic neurons. J Neu-rosci 14:740–750.

’Malley JP, Moore CT, Salpeter MM (1997) Stabilization of acetyl-choline receptors by exogenous ATP and its reversal by cAMP andcalcium. J Cell Biol 138:159–165.

atton BL, Cunningham JM, Thyboll J, Kortesmaa J, Westerblad H,Edstrom L, Tryggvason K, Sanes JR (2001) Properly formed butimproperly localized synaptic specializations in the absence oflaminin �4. Nat Neurosci 4:597–604.

alevic V, Burnstock G (1998) Receptors for purines and pyrimidines.Pharmacol Rev 50:413–492.

edman RS, Silinsky EM (1994) ATP released together with acetyl-choline as the mediator of neuromuscular depression at frog motornerve endings. J Physiol 477:117–127.

obitaille R (1995) Purinergic receptors and their activation by endog-enous purines at perisynaptic glial cells of the frog neuromuscularjunction. J Neurosci 15:7121–7131.

uppelt A, Ma W, Borchardt K, Silberberg SD, Soto F (2001)Genomic structure, developmental distribution and functionalproperties of the chicken P2X receptor. J Neurochem 77:
51256 –1265.
yten M, Hoebertz A, Burnstock G (2001) Sequential expression ofthree receptor subtypes for extracellular ATP in developing ratskeletal muscle. Dev Dyn 221:331–341.

yten M, Dunn PM, Neary JT, Burnstock G (2002) ATP regulates thedifferentiation of mammalian skeletal muscle by activation of aP2X5 receptor on satellite cells. J Cell Biol 158:345–355.

yten M, Yang SY, Dunn PM, Goldspink G, Burnstock G (2004)Purinoceptor expression in regenerating skeletal muscle in themdx mouse model of muscular dystrophy and in satellite cellcultures. FASEB J 18:1404–1406.

anes JR, Lichtman JW (1999) Development of the vertebrate neuro-muscular junction. Annu Rev Neurosci 22:389–442.

anes JR, Lichtman JW (2001) Induction, assembly, maturation andmaintenance of a postsynaptic apparatus. Nat Rev Neurosci2:791–805.

earl TJ, Redman RS, Silinsky EM (1998) Mutual occlusion of P2XATP receptors and nicotinic receptors on sympathetic neurons ofthe guinea-pig. J Physiol (Lond) 510:783–791.

ilinsky EM, Redman RS (1996) Synchronous release of ATP andneurotransmitter within milliseconds of a motor nerve impulse inthe frog. J Physiol 492:815–822.

later CR (2003) Structural determinants of the reliability of synaptictransmission at the vertebrate neuromuscular junction. J Neurocy-tol 32:505–522.

homas SA, Zawisa MJ, Lin X, Hume RI (1991) A receptor that ishighly specific for extracellular ATP in developing chick skeletalmuscle in vitro. Br J Pharmacol 1034:1963–1969.

ells DG, Zawisa MJ, Hume RI (1995) Changes in responsiveness toextracellular ATP in chick skeletal muscle during development andupon denervation. Dev Biol 172:585–590.

hou X, Galligan JJ (1998) Non-additive interaction between nicotiniccholinergic and P2X purine receptors in guinea-pig enteric neurons

in culture. J Physiol 513:685–697.

(Accepted 5 July 2007)(Available online 17 July 2007)

ABNORMALITIES IN NEUROMUSCULAR JUNCTION STRUCTURE AND SKELETAL MUSCLE FUNCTION IN MICE LACKING THE P2X2 NUCLEOTIDE RECEPTOREXPERIMENTAL PROCEDURESAnimalsImmunohistochemistry and histologyAnalysis of AChR expression at NMJsMeasurement of muscle fiber numberAssessment of the incidence of centrally nucleated fibersMeasurement of muscle fiber cross-sectional areaElectron microscopyPhysiologyNerve-mediated and muscle-mediated twitch responsesTubocurarine sensitivitySingle channel recording from muscle fiber end-plates

RESULTSP2X2 receptor expression in developing skeletal muscleAbnormalities in NMJ structure in P2X2 / skeletal muscle visualized by light microscopyAbnormalities in the ultrastructure of NMJs in P2X2 / soleus muscleChannel properties of AChRs expressed in P2X2 receptor positive wild-type and P2X2 / muscle fibersMuscle fiber atrophy in soleus muscle from P2X2 / miceChanges in response to tetanic stimulation in P2X2 / soleus muscleFast (type II) muscle fibers in wild-type and P2X2 / soleus muscle

DISCUSSIONCONCLUSIONREFERENCES

abnormalities in neuromuscular junction structure …s copies/cv1357.pdf · 2010. 9. 3. ·...

Documents